skills/razbakov/skills/weekly-review/Gen Agent Trust Hub

weekly-review

Pass

Audited by Gen Agent Trust Hub on Jun 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands via the gog cal CLI utility to read and write events to the user's Google Calendar.
  • [PROMPT_INJECTION]: The skill ingests untrusted data from sessions/ (browser history and AI logs) and project README.md files. This creates an indirect prompt injection surface where malicious instructions embedded in these logs or files could influence the agent's behavior during the review process.
  • Ingestion points: Reads sessions/ folder and project README.md files.
  • Boundary markers: None explicitly defined for interpolated data.
  • Capability inventory: Executes gog cal commands and writes to local markdown files (PROJECTS.md, now.md, reviews/).
  • Sanitization: No visible sanitization of ingested content before it is processed by the AI.
  • [DATA_EXFILTRATION]: The skill accesses sensitive local information including browser history and contact commitments. While it sends data to Google Calendar, this is a core functionality of the skill requested by the user, and no exfiltration to unauthorized third-party domains was detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 24, 2026, 06:30 AM
Security Audit — agent-trust-hub — weekly-review